Day 3 morphology is a poor predictor ofblastocyst quality in extended culture

James Graham, M.S., Taer Han, B.S., Richard Porter, Michael Levy, M.D.,Robert Stillman, M.D., and Michael J. Tucker, Ph.D.

Shady Grove Reproductive Science Center, Rockville, Maryland

Objective: To determine how the quality of blastocysts formed on day 5/6 of extended culture compares withtheir morphology on day 3.

Design: Retrospective observational study of IVF laboratory records.

Setting: Private assisted reproduction clinic.

Patient(s): 101 IVF cycles in which 5 to 25 embryos were produced. The average maternal age was 33.1years.

Intervention(s): Embryos were individually cultured in vitro in sequential media for an extended time toenable use of blastocysts for fresh transfer or cryopreservation.

Main Outcome Measure(s): Comparison of embryo quality for putative ET or cryopreservation on day 3with quality of embryos used for actual ET and cryopreservation on day 5/6.

Result(s): Of 1,263 cleaving embryos, 559 were judged to have been suitable for use on day 3; 355 wouldhave been used for ET (average per ET, 3.5) and 204 would have been frozen (equivalent to 44% utilization).In actuality, 471 blastocysts were used on day 5/6, of which 234 were transferred (average per ET, 2.3), and237 were frozen (equivalent to 37% utilization). Only 48% embryos that would have been chosen for ETand/or cryopreservation on day 3 were eventually used in such a manner at the blastocyst stage. Historically,the rate of viable pregnancy from day 3 transfers was 30.5% per transfer; this rate increased to 45% withroutine day 5/6 transfers.

Conclusion(s): Extended culture of human embryos seems to increase discrimination of potential embryonicviability. Criteria for embryo selection on day 3 seem to be inadequate. Extended in vitro culture may thereforebe an effective means of optimizing IVF clinical success. (Fertil Sterilt 2000;74:495–7. ©2000 by AmericanSociety for Reproductive Medicine.)

Key Words: Extended culture, blastocyst transfer, embryo quality, morphology

The need to culture human embryos forlonger periods of their development is becom-ing increasingly clear. Extended in vitro cultureincreases the degree of embryo selection, thusmaximizing the choice of the best quality em-bryos for transfer and minimizing the need fortransfer of greater numbers of embryos. Inthis manner, IVF success may be optimizedthrough increased pregnancy rates and the riskfor multiple implantation may be reduced (1).

The simple method of extending embryoculture from 2 to 3 days has improved overallclinical IVF results in the past decade. It seemsthat the time has arrived for the culture periodto be extended to day 5/6, thus allowing em-bryos to “select” themselves by growth to theblastocyst stage. However, blastocyst forma-tion itself does not fully reflect the viability of

the embryo (1), and not all blastocysts are ofequal quality. Experience in defining betterblastocyst quality in terms of morphology anddevelopmental rate is growing. Close attentionto the cell count in the trophectoderm and innercell mass, rate of blastocyst formation, andhistorical data on pronuclear and cleavagestage status in each embryo are all useful indiscerning late-stage preimplantation embryoquality.

To investigate the relative merits of embryoselection on day 3 compared with day 5/6, weperformed a retrospective comparison at ourcenter. Embryos were grown individually tothe blastocyst stage and were scored on day 3,the traditional time of potential selection, andon day 5/6, when embryos were actually se-lected for ET and cryopreservation.

Received January 1, 2000;accepted March 8, 2000.Presented in part at the55th Annual Meeting of theAmerican Society forReproductive Medicine,Toronto, Ontario, Canada,September 25–30, 1999.Reprint requests: MichaelJ. Tucker, Ph.D., GeorgiaReproductive Specialists,5445 Meridian Mark Road,Suite 270, Atlanta, Georgia30348 (FAX: 404-843-0812;E-mail: mtucker@ivf.com).

FERTILITY AND STERILITY tVOL. 74, NO. 3, SEPTEMBER 2000Copyright ©2000 American Society for Reproductive MedicinePublished by Elsevier Science Inc.Printed on acid-free paper in U.S.A.

0015-0282/00/$20.00PII S0015-0282(00)00689-0

495

MATERIALS AND METHODS

We undertook in vitro culture in 101 IVF cycles in whichwomen (average age, 33.1 years [range, 24–44 years]) hadsufficient embryos (range, 5 to 25) to justify some degree ofembryo selection. Oocytes were collected and placed intohuman tubal fluid medium plus human serum albumin (InVitro Care, San Diego, CA). Thirty-five–mL droplets ofmedium were placed under mineral oil in 60-mm culturedishes (Nunc, Denmark). Insemination was either carried outconventionally or by ICSI. Zygotes were placed into indi-vidually marked 15-mL droplets of medium under mineraloil in 35-mm culture dishes (Nunc).

For initial (day 1 to day 3) and extended culture (day 3 today 5/6), either Basal XI (In Vitro Care) plus S-2 (IVFScience, Gothenburg, Sweden) media (n5 91) or G1 plusG2 media (IVF Science) (n5 10) were used in a sequentialmanner. Embryos were maintained in 5.5% CO2 in air up today 3 and subsequently in 5% CO2 in air for extended cul-ture. Blastocyst development was expected on day 5 or 6; anestimate of when development might occur was based ongrowth over the first few days. No apparent differences in em-bryonic development were observed in the two media groups.

Embryo development was judged daily; traditional as-sessment for ET and cryopreservation was made on day 3 byusing the traditional criteria for selection: morphology (prin-cipally regularity of the blastomeres and fragmentation), rateof development, and cell nucleation (2). All embryos weresubsequently grown until day 5 or 6, when they were againscored for actual ET and cryopreservation. For each embryo,we compared directly its traditional selection potential onday 3 with its ultimate use or nonuse day 5/6.

Because no interventions were used in this retrospectiveobservational study, no Institutional Review Board approvalwas required.

RESULTS

Of 1,263 cleaving embryos, 559 were judged to be ofsufficient quality for use on day 3 on the basis of traditionalselection criteria. Day 3 use would have resulted in transferof 355 embryos, assuming an average of 3.5 embryos per ET(the average number of embryos transferred on day 3 in ourprogram during the study period). The remaining 204 em-bryos would have been cryopreserved. The total potentialutilization rate was therefore 44%. However, in reality, only234 blastocysts were transferred on day 5/6 (an average of2.3 embryos per ET) and 237 embryos were cryopreserved;the embryo utilization rate at this stage was therefore 37%(Table 1, Fig. 1).

Extended culture resulted in a 16% decrease in usableembryos from day 3 to day 5/6 (471 compared with 559).Only 48% (265 of 559) of the embryos that would have beenconsidered suitable for ET or cryopreservation on day 3 were

chosen for ET or cryopreservation at the blastocyst stage onday 5/6.

A historical comparison of the rates of viable pregnancyfrom embryo transfer on day 3 using the culture conditionsdescribed here, before extended culture to day 5/6 wasinstituted, indicates a distinct improvement. The pregnancyrate increased from 30.5% to 45% per ET after embryoswere grown to the blastocyst stage, and more important, theembryonic implantation rate increased from 16.5% to 24%.

DISCUSSION

Extended in vitro culture of human embryos to day 5/6seems to increase discrimination of potential embryonic vi-ability with minimal deleterious effect. However, in agree-ment with results of a previous study (3), traditional criteriafor embryo selection on day 3 for ET seem to be relatively

T A B L E 1

Comparison of potential embryo use on day 3 with actualuse on day 5/6.

Variable Day 3 Day 5/6Selected on

both occasions

No. of embryos transferred 355 234 143No. of embryos cryopreserved 204 237 122Total 559 471 265

Graham. Day 3 morphology is a poor predictor. Fertil Steril 2000.

F I G U R E 1

Predicted embryo use on day 3 compared with actual use onday 5/6. Black bars indicate embryos that would have beenselected on day 3; white bars indicate embryos selected onday 5/6 that were not selected on day 3.

Graham. Day 3 morphology is a poor predictor. Fertil Steril 2000.

496 Graham et al. Embryo utilization on day 3 versus day 5/6 Vol. 74, No. 3, September 2000

inefficient. We cultured individually all embryos from thezygote stage onward so that historical data on each embryocould be used to ascertain the ultimate use of that embryo(fresh transfer or cryopreservation).

On day 3, the selection potential for both transfer andcryopreservation at the cleavage stage seem to be inade-quate. In our study, only 48% of embryos chosen for use onday 3 were chosen again on day 5/6. This result is similar tothe rate of 51% found by Rijnders and Jansen (3), whoconsidered only fresh transfer.

This improvement in embryo discrimination by extendedculture does not automatically translate to better pregnancyrates, though our limited historic comparison suggests that itmight. Pregnancy rates do not necessarily increase withblastocyst transfers; day 3 transfers usually involve the useof more embryos, which compensates for the poorer selec-tion potential at that stage compared with transfer at theblastocyst stage, in which fewer embryos are usually trans-ferred.

Various factors affect blastocyst formation and quality,including culture conditions, number of oocytes, maternalage, and male factor infertility (1). In addition, the quality ofearly-stage embryos can substantially influence rates of blas-tocyst formation rates (4–6). However, our data do nottotally support the idea that the number of eight-cell embryoson day 3 and the potential for blastocyst formation aredirectly correlated (5); in particular, they do not support theassumption that the blastocysts that form do so from theday-3 eight-cell embryos. Confirmation of this assumption isonly possible if all embryos are individually cultured.

Our study compared quality of individual day-3 embryoswith their use as blastocysts and found a lack of correlationbetween quality at both stages in about half of the embryos.Seemingly unsuitable cleavage-stage embryos can developto the blastocyst stage and beyond to produce healthy preg-nancies (1). A reasonable compromise to make routine IVFtherapy more efficient while continuing to perform day 3transfers would be to place all unused embryos in extendedculture and cryopreserve them only at the blastocyst stage.This approach does not capitalize on the seemingly im-proved embryo discrimination with day 5/6 transfer, but itdoes reduce the number of potentially nonviable embryosthat are cryopreserved.

In our study, embryo use decreased with extended cul-ture; 44% of embryos would have been transferred andfrozen on day 3 compared with 37% on day 5/6. However,this decrease is not large, and given the increased selectionpotential for embryos on day 5/6, extended culture seemsjustified. The average day of blastocyst transfer was day 5.4.By extending culture to day 6 before transfer, it was possibleto increase the number of blastocysts to choose from; insome cases no early cavitating blastocysts were seen on day5 to choose for transfer. Speed of development to the blas-tocyst stage and potential for pregnancy are clearly corre-lated; however, use of assisted hatching procedures onslower-developing embryos seems to compensate for thedifference (7). Many of these blastocysts may have arisenfrom what would have appeared to be suboptimal embryos atthe cleavage stage, based on traditional selection criteria.

Acknowledgements:The authors thank the following persons for theirclinical and technical skills during this study: Stephen Greenhouse, M.D.,Gil Mottla, M.D., Art Sagoskin, M.D., Eric Widra, M.D., Paula Morton,B.S., and Cindy Sweitzer, B.S.

REFERENCES1. Jones GM, Trounson AO. Blastocyst stage transfer: pitfalls and bene-

fits. Hum Reprod 1999;14:1405–8.2. Veeck LL. An atlas of human gametes and conceptuses. New York:

Parthenon, 1999.3. Rijnders PM, Jansen CAM. The predictive value of day 3 embryo

morphology regarding blastocyst formation, pregnancy and implanta-tion rate after day 5 transfer following in-vitro fertilization or intracy-toplasmic sperm injection. Hum Reprod 1998;13:2869–73.

4. Miller KF, Fry KL, Arciaga RL, Attaran M, Goldberg JM, Falcone T.Correspondence of day 3 selection of best quality embryos and actualdevelopment to blastocyst stage on day 5 (abstract). Hum Reprod1999;14(Suppl 1):187–8.

5. Racowsky C, Jackson KV, Cekleniak NA, Fox JH, Hornstein MD,Ginsburg ES. The number of eight-cell embryos is a key determinantfor selecting day 3 or 5 transfer. Fertil Steril 2000;73:558–64.

6. Balaban B, Isiklar A, Aksoy S, Alatas C, Mercan R, Nuhoglu A, et al.The effect of day 3 embryo quality on blastocyst formation and quality(abstract). Proceedings of the Annual Meeting of the American Societyfor Reproductive Medicine, Toronto, Ontario, Canada, September 25–30, 1999. S110.

7. Tucker M. Relevance of assisted hatching with blastocyst stage trans-fer. Proceedings of the First World Congress on Controversies inObstetrics Gynecology & Infertility, Prague, Czech Republic, October28–31, 1999. Bologna, Italy: Monduzzi, 1999:49–52.

FERTILITY & STERILITY t 497